Is it possible to shoot bad dark frames?

[Photosbykev]

The screen print below is of two images containing identical lights, flats and bias frames. The right hand one has a fresh set of Darks added.

The 240 second darks were shot at the same iso, exposure and temperature on a stock Canon 5D4

I've process the two images with exactly the same PixInsight settings (just background neutralisation,  colour calibration and stf applied to see them) DBE wasn't applied because that would affect the output for this test.

I'm getting these strange slightly curved bands, ringed in the right hand image when I add the darks during the calibration routine

Is it possible to shoot bad dark frames? Previous sets of darks on the same setup have worked correctly so I'm struggling to see why this dark set would be 'faulty'?

The same effect is seen on other data sets as well using this set of darks.

For reference this is what I get with a basic edit from the lights/flats/bias image

regards

Kev

Edited December 4, 2019 by Photosbykev

[cjdawson]

Yep, it's possible to shoot bad darks.  Light leakage would be the first culprit for that.   For a DSLR, you have two places where light can leak into the sensor, first is obviously the lens.  The second less obvious is the viewfinder.

When shooting darks, it's important to make sure that both as properly covered and are not leaking light to ensure that dark frames really are dark, without other influences.   Putting the dSLR in a box, then putting that in a cupboard usually works, however be aware that the temperature of the dark will therefore change, as the camera will get warm and may not have the ventilation needed to keep the temperature down.

[andrew s]

I always take darks in the dark just to be sure. Light leaks can also be an issue for flats. 

Regards Andrew 

[Photosbykev]

16 minutes ago, cjdawson said:

Yep, it's possible to shoot bad darks.  Light leakage would be the first culprit for that.   For a DSLR, you have two places where light can leak into the sensor, first is obviously the lens.  The second less obvious is the viewfinder.

When shooting darks, it's important to make sure that both as properly covered and are not leaking light to ensure that dark frames really are dark, without other influences.   Putting the dSLR in a box, then putting that in a cupboard usually works, however be aware that the temperature of the dark will therefore change, as the camera will get warm and may not have the ventilation needed to keep the temperature down.

Thank you for the input CJ

My darks are normally very good as I do cover the refractor with it's dust cover and then slide a black fabric bag over the end to make sure the objective end of the refractor is  completely blacked out and the camera viewfinder eyepiece is covered with a custom eyepiece blank. The darks and lights sensor temperature were within 2degreesC of each other so I'm struggling with why they should be wrong. I can reshoot them, and probably will, one evening this week when the temperatures get down to around zero ambient and check the second dark set.

regards

Kev

[Photosbykev]

1 minute ago, andrew s said:

I always take darks in the dark just to be sure. Light leaks can also be an issue for flats. 

Regards Andrew 

I shoot the flats immediately after the imaging session with a battery powered flat field panel so it's very dark lol. I think the flats are ok looking at them in PI, certainly not significantly different to other flats I've taken. It's only when I add the darks to the calibration stack that the image gets the curved banding across it in the lower right corner. It isn't the standard Canon sensor banding issue which I normally correct in PI anyway

regards

Kev

[reddoss]

1 minute ago, Photosbykev said:

Thank you for the input CJ

My darks are normally very good as I do cover the refractor with it's dust cover and then slide a black fabric bag over the end to make sure the objective end of the refractor is  completely blacked out and the camera viewfinder eyepiece is covered with a custom eyepiece blank. The darks and lights sensor temperature were within 2degreesC of each other so I'm struggling with why they should be wrong. I can reshoot them, and probably will, one evening this week when the temperatures get down to around zero ambient and check the second dark set.

regards

Kev

When I was using a DSLR for astro-imaging sometimes at the end of a long night I would forget to take darks. As a quick fix I would take darks by putting the camera in the fridge at home to get close to the outside temperature as possible.

[Photosbykev]

Problem solved but I don't know the cause. Two of the thirty dark images look like this heavily stretched!! New darks required lol

Thanks for all the input folks

Kev

Edited December 4, 2019 by Photosbykev

[cjdawson]

light leakage or amp glow maybe?

[Photosbykev]

33 minutes ago, cjdawson said:

light leakage or amp glow maybe?

Light leakage doesn't make sense as all the other darks are fine. The canon sensor does have bias noise along the bottom edge but no obvious amp glow. It does look like light leakage but how I'm not sure. I'll reshoot them and check the individual files

Kev